测量与量表：基本原理和比较量表.pptx

ChapterEightMeasurementAndScaling:FundamentalsAndComparativeScaling第8章测量与量表：基本原理和比较量表,赵冬阳讲师经济与管理学院市场营销博士生e-mail:sunny_cn1206,ChapterObjectives本章目标,1.Introducetheconceptsofmeasurementandscalingandshowhowscalingmaybeconsideredanextensionofmeasurement.2.Explainthecharacteristicsofdescription,order,distance,andoriginandhowtheydefinethelevelofmeasurementofascale.3.Discusstheprimaryscalesofmeasurementanddifferentiatenominal,ordinal,interval,andratioscales.4.Classifyanddiscussscalingtechniquesascomparativeandnon-comparative,anddescribethecomparativetechniquesofpairedcomparison,rankorder,constantsum,andQ-sortscaling.,1理解测量和量表的概念，并说明如何将量表视为测量的一个扩充。2.解释描述、顺序、距离和原点的特征，和如何定义量表测量尺度3.讨论测量的主要尺度并区分定类、定序、定距和定比尺度。4.讨论并将量表技术分类比较和非比较量表，描述比较量表中的配对比较量表、等级顺序量表、常量量表和Q分类量表。,2011/2/17,ZHAODongyangPanzhihuaUniversity,2,ChapterOutline章节提纲,1)Overview2)MeasurementandScaling3)ScalesCharacteristicsDescriptionOrderDistanceOrigin,概要测量和标度尺度的特征描述比较程度原点,2011/2/17,ZHAODongyangPanzhihuaUniversity,3,ChapterOutline,4)PrimaryScalesofMeasurementNominalScaleOrdinalScaleIntervalScaleRatioScale5)AComparisonofScalingTechniques,主要的测量尺度定类尺度定序尺度定距尺度定比尺度量表技术的比较,2011/2/17,ZHAODongyangPanzhihuaUniversity,4,ChapterOutline,6)ComparativeScalingTechniquesPairedComparisonRankOrderScalingConstantSumScalingQ-SortandOtherProcedures7)VerbalProtocols8)InternationalMarketingResearch9)EthicsinMarketingResearch10)Summary,比较量表技术配对比较等级顺序量表常量和量表Q分类和其他方法,2011/2/17,ZHAODongyangPanzhihuaUniversity,5,MeasurementandScaling测量和标度,Measurementmeansassigningnumbersorothersymbolstocharacteristicsofobjectsaccordingtocertainpre-specifiedrules.One-to-onecorrespondencebetweenthenumbersandthecharacteristicsbeingmeasured.Therulesforassigningnumbersshouldbestandardizedandapplieduniformly.Rulesmustnotchangeoverobjectsortime.,测量意味着根据某些预先指定的规则配给对象的特征分配数字或其它符号。数字和被测的特点一一对应关系。编号分配规则应标准化和统一使用。规则不能对象或时间更改。,2011/2/17,ZHAODongyangPanzhihuaUniversity,6,ScaleCharacteristics尺度特征,2011/2/17,ZHAODongyangPanzhihuaUniversity,7,ScaleCharacteristics,2011/2/17,ZHAODongyangPanzhihuaUniversity,8,量表尺度上的不同特征,2011/2/17,ZHAODongyangPanzhihuaUniversity,9,MeasurementandScaling测量和标度,Scalinginvolvescreatingacontinuumuponwhichmeasuredobjectsarelocated.Consideranattitudescalefrom1to100.Eachrespondentisassignedanumberfrom1to100,with1=ExtremelyUnfavorable,and100=ExtremelyFavorable.Measurementistheactualassignmentofanumberfrom1to100toeachrespondent.Scalingistheprocessofplacingtherespondentsonacontinuumwithrespecttotheirattitudetowarddepartmentstores.,标度涉及产生一个将被测物体连续区间上的定位。考虑态度的尺度从1到100。每名回答都指定一个从1到100的编号。1=极为不喜欢和100=极为喜欢。测量是每个回答的介于1到100实际分配。标度是根据他们对百货公司的态度在连续区间定位回答的过程。,2011/2/17,ZHAODongyangPanzhihuaUniversity,10,PrimaryScalesofMeasurement,2011/2/17,ZHAODongyangPanzhihuaUniversity,11,PrimaryScalesofMeasurement：NominalScale测量的主尺度：定类尺度,Thenumbersserveonlyaslabelsortagsforidentifyingandclassifyingobjects.Whenusedforidentification,thereisastrictone-to-onecorrespondencebetweenthenumbersandtheobjects.Thenumbersdonotreflecttheamountofthecharacteristicpossessedbytheobjects.Theonlypermissibleoperationonthenumbersinanominalscaleiscounting.Onlyalimitednumberofstatistics,allofwhicharebasedonfrequencycounts,arepermissible,e.g.,percentages,andmode.,数字只作为用于对象的识别和分类的标签或标志。用于识别时,数字和对象之间有一一的对应。数字并不反映对象所拥有的特征的数量。只允许对定类尺度中的数字进行计数操作。只允许计算少数的基于频率计数的统计量，例如，百分比、众数、卡方和二项式检验。,2011/2/17,ZHAODongyangPanzhihuaUniversity,12,用于不同尺度的问题例子(1),13,2011/2/17,ZHAODongyangPanzhihuaUniversity,PrimaryScalesofMeasurement：OrdinalScale测量的主要尺度：定序尺度,Arankingscaleinwhichnumbersareassignedtoobjectstoindicatetherelativeextenttowhichtheobjectspossesssomecharacteristic.Candeterminewhetheranobjecthasmoreorlessofacharacteristicthansomeotherobject,butnothowmuchmoreorless.Anyseriesofnumberscanbeassignedthatpreservestheorderedrelationshipsbetweentheobjects.Inadditiontothecountingoperationallowablefornominalscaledata,ordinalscalespermittheuseofstatisticsbasedoncentiles,e.g.,percentile,quartile,median.,2011/2/17,ZHAODongyangPanzhihuaUniversity,14,用于不同尺度的问题例子(2),15,2011/2/17,ZHAODongyangPanzhihuaUniversity,PrimaryScalesofMeasurement:IntervalScale测量的主要尺度：定距尺度,Numericallyequaldistancesonthescalerepresentequalvaluesinthecharacteristicbeingmeasured.Itpermitscomparisonofthedifferencesbetweenobjects.Thelocationofthezeropointisnotfixed.Boththezeropointandtheunitsofmeasurementarearbitrary.Anypositivelineartransformationoftheformy=a+bxwillpreservethepropertiesofthescale.Itisnotmeaningfultotakeratiosofscalevalues.Statisticaltechniquesthatmaybeusedincludeallofthosethatcanbeappliedtonominalandordinaldata,andinadditionthearithmeticmean,standarddeviation,andotherstatisticscommonlyusedinmarketingresearch.,2011/2/17,ZHAODongyangPanzhihuaUniversity,16,用于不同尺度的问题例子(3),17,定距量表的问题1.请根据总体偏好划分以下品牌的级别2.用圆圈标出你对以下几句话的同意程度：,2011/2/17,ZHAODongyangPanzhihuaUniversity,PrimaryScalesofMeasurement：RatioScale测量的主要尺度：定比尺度,2011/2/17,ZHAODongyangPanzhihuaUniversity,18,用于不同量表的问题例子(3),19,2011/2/17,ZHAODongyangPanzhihuaUniversity,IllustrationofPrimaryScalesofMeasurement测量的主要尺度的说明Table8.1,2011/2/17,ZHAODongyangPanzhihuaUniversity,20,PrimaryScalesofMeasurement,Table8.2,2011/2/17,ZHAODongyangPanzhihuaUniversity,21,PrimaryScalesofMeasurement主要的测量量表,Table8.2,2011/2/17,ZHAODongyangPanzhihuaUniversity,22,AClassificationofScalingTechniques量表技术的分类,Figure8.2,2011/2/17,ZHAODongyangPanzhihuaUniversity,23,AComparisonofScalingTechniques,2011/2/17,ZHAODongyangPanzhihuaUniversity,24,AComparisonofScalingTechniques,2011/2/17,ZHAODongyangPanzhihuaUniversity,25,RelativeAdvantagesofComparativeScales比较量表的相对优点,2011/2/17,ZHAODongyangPanzhihuaUniversity,26,Smalldifferencesbetweenstimulusobjectscanbedetected.Sameknownreferencepointsforallrespondents.Easilyunderstoodandcanbeapplied.Involvefewertheoreticalassumptions.Tendtoreducehaloorcarryovereffectsfromonejudgmenttoanother.,RelativeDisadvantagesofComparativeScales相对缺点,2011/2/17,ZHAODongyangPanzhihuaUniversity,27,OrdinalnatureofthedataInabilitytogeneralizebeyondthestimulusobjectsscaled.,定序性质无法推广到被计量的刺激物体之外。,ComparativeScalingTechniquesPairedComparisonScaling比较量表技术-配对比较量表,2011/2/17,ZHAODongyangPanzhihuaUniversity,28,ObtainingShampooPreferencesUsingPairedComparisons用配对比较法获得香波偏好的信息,2011/2/17,ZHAODongyangPanzhihuaUniversity,29,PairedComparisonSelling,Themostcommonmethodoftastetestingispairedcomparison.Theconsumerisaskedtosampletwodifferentproductsandselecttheonewiththemostappealingtaste.Thetestisdoneinprivateandaminimumof1,000responsesisconsideredanadequatesample.Ablindtastetestforasoftdrink,whereimagery,self-perceptionandbrandreputationareveryimportantfactorsintheconsumerspurchasingdecision,maynotbeagoodindicatorofperformanceinthemarketplace.TheintroductionofNewCokeillustratesthispoint.NewCokewasheavilyfavoredinblindpairedcomparisontastetests,butitsintroductionwaslessthansuccessful,becauseimageplaysamajorroleinthepurchaseofCoke.Apairedcomparisontastetest,2011/2/17,ZHAODongyangPanzhihuaUniversity,30,ComparativeScalingTechniquesRankOrderScaling比较量表技术等级顺序量表,2011/2/17,ZHAODongyangPanzhihuaUniversity,31,PreferenceforToothpasteBrandsUsingRankOrderScaling使用等级顺序量表获取牙膏品牌的偏好信息,2011/2/17,ZHAODongyangPanzhihuaUniversity,32,PreferenceforToothpasteBrandsUsingRankOrderScaling使用等级顺序量表获取牙膏品牌的偏好信息,2011/2/17,ZHAODongyangPanzhihuaUniversity,33,ComparativeScalingTechniquesConstantSumScaling常量和量表,2011/2/17,ZHAODongyangPanzhihuaUniversity,34,ImportanceofBathingSoapAttributesUsingaConstantSumScale用常量和量表统计浴皂属性的重要性,2011/2/17,ZHAODongyangPanzhihuaUniversity,35,ImportanceofBathingSoapAttributesUsingaConstantSumScale,2011/2/17,ZHAODongyangPanzhihuaUniversity,36,2011/2/17,ZHAODongyangPanzhihuaUniversity,37,Allrightsreserved.Nopartofthispublicationmaybereproduced,storedinaretrievalsystem,ortransmitted,inanyformorbyanymeans,electronic,mechanical,photocopying,recording,orotherwise,withoutthepriorwrittenpermissionofthepublisher.PrintedintheUnitedStatesofAmerica.Copyright2010PearsonEducation,Inc.publishingasPrenticeHall,